Patients are defined as internal medicine patients when more than one system is involved. Such patients are lacking in the clinical and epidemiological literature. Among 330 patients discharged from a unit of internal medicine, 212 (64%) were prospectively classified in that category. As compared to the "specialty patients", the "internal medicine patients" were almost the same age (56 vs 53 years) and sex, but more diagnoses were recorded for each: 4.16 vs 2.9 p less than 0.001; their death rate was apparently higher: 12.7% vs 7.6% (NS). Therapeutic results were poor. Psychic disturbances were prevailing, with classical mental diseases, but also major somatic consequences of seemingly minor behavioral deviations. This morbidity is largely avoidable. The discussion addresses the validity of results and their confrontation with data collated in the literature; a few specific characteristics of care; the widely variable psychological reactions of nurses, students, residents and physicians; risks of rejection from the general hospital, for technical and economic reasons; and lastly, risks of rejection by public opinion. Who is prepared to be involved with major morbidity which is time-consuming, largely refractory to management, preventable, unchallenging for some though instructive for others, unsaleable and useless as a vote-catcher?